Reward tiers pay at 90%; the top tier needs 2 more taps

You've seen the screen. A progress bar, three tiers, the first one already lit up. "You're 90% of the way to your next reward." And then, just below it, a smaller line: two more qualifying purchases to reach the top tier. The math doesn't quite add up in your head, but your thumb is already moving. That tiny gap between where the interface says you are and where the fine print says you need to be is not a design flaw. It's the whole point.

The two-tap gap is doing more work than the reward itself

Card issuers and payment networks have gotten very good at something that has nothing to do with interest rates or interchange fees: they've gotten good at designing the last mile of a reward journey.

Consider the structure. A program tells you that you've unlocked 90% of the value. That's a strong signal — you're almost there, the effort is nearly sunk, quitting now feels like wasting what you've already earned. Then the same screen quietly reveals that the final tier requires two more transactions. Not one, which might feel trivial. Not five, which might feel like a chore. Two. It's a number small enough to feel achievable and large enough to require a deliberate decision.

That deliberate decision is the product. The reward is just the excuse.

Behavioral economists have a name for the discomfort you feel at 90%: it's a close cousin of loss aversion, the tendency — documented extensively by Kahneman and Tversky — to weigh losses more heavily than equivalent gains. Once a reward feels "yours," not completing the journey registers as a loss, not a missed gain. The interface manufactures ownership before you've actually earned anything, and then charges you two taps to keep it.

Why "almost there" beats "here's a discount"

There's a reason issuers rarely lead with a flat cashback rate anymore. A flat 1.5% is transparent, predictable, and emotionally inert. It doesn't create a loop.

A tiered structure does. And the specific mechanic at work is a close relative of variable-ratio reinforcement — the pattern where a reward arrives after an unpredictable number of actions, which produces more persistent behavior than a fixed, guaranteed payout. You don't know exactly which purchase will tip you into the next tier, so you keep checking, keep spending, keep opening the app. The progress bar updates. Something in your brain registers a small hit of satisfaction. The loop closes, and then reopens one tier higher.

What's clever about the "90% plus two more taps" construction is that it combines two reinforcing patterns at once. The near-completion triggers the goal-gradient effect — the well-documented finding that people accelerate their effort as they approach a goal, first described by Clark Hull in the 1930s and later applied to everything from loyalty cards to coffee shop stamps. Meanwhile, the small remaining requirement keeps the loop from ever fully resolving. You're always 90% done. You're never done.

The study that should make you look twice

The coffee stamp card is the classic version of this. Researchers ran a now-famous field experiment where customers received either a card requiring 8 stamps for a free drink or a card requiring 10 stamps — but with 2 stamps already filled in. Same effective requirement: 8 purchases. The group with the pre-filled card, the ones who started at "20% done," completed the card significantly faster.

Same effort. Different framing. Different behavior.

Reward tiers in banking work on the identical principle, just with more decimal places. When an app tells you that you've already earned 90% of a tier's value, it's pre-filling your stamp card. When it tells you two more taps will finish the job, it's setting the finish line just far enough away to feel meaningful and just close enough to feel inevitable.

The uncomfortable implication is that the reward itself — the lounge access, the higher cashback rate, the bonus points — matters less than the structure of the chase. Programs that pay out generously but present the payout as a flat, finished thing tend to underperform programs that pay out less but keep the bar moving.

Where the design crosses from clever to costly

Here's where it gets genuinely interesting, and where the payment industry is starting to feel some friction.

The same mechanics that drive engagement also drive a specific kind of decision-making under uncertainty. You don't know, at the moment of the second tap, whether the purchase you're about to make is one you'd have made anyway. The tier structure nudges you toward spending that serves the progress bar rather than your actual budget. That's not a bug from the issuer's perspective. It is, however, a real cost from yours.

And there's a growing regulatory spotlight on this. Several jurisdictions have started scrutinizing whether "nearly there" messaging in financial apps constitutes a misleading representation of value. The argument isn't that rewards are bad. It's that a progress bar showing 90% when the actual qualifying threshold is meaningfully higher can mislead a reasonable consumer about what they've actually earned.

The industry's defense — that the terms are disclosed, that the tiers are clearly defined — runs into a well-known problem. Disclosure doesn't undo framing. If it did, loss aversion wouldn't exist, and neither would the entire field of behavioral economics. People don't process fine print the way they process a glowing progress bar. That asymmetry is the entire business model.

What the next version of this looks like

The interesting question isn't whether reward tiers will keep using the two-tap gap. They will. The question is what happens when the gap gets personalized.

We're already seeing early versions. Some programs now adjust the remaining-requirement messaging based on your spending pattern — if you're a heavy user, the final tier might require four more transactions instead of two, because the system has learned you'll complete it anyway. If you're a light user, the gap shrinks to keep you from disengaging. The reward structure isn't a fixed ladder anymore. It's a dial.

That's a meaningful shift, and it cuts both ways. A well-tuned dial can be genuinely useful — it can surface a reward that's actually within reach for your spending habits, rather than dangling an unreachable tier that just frustrates you. A badly tuned one becomes a machine for manufacturing near-misses, each one calibrated to the exact point where you're most likely to tap again.

The practical move, if you're on the consumer side of this, is to notice the moment the progress bar starts feeling like an obligation. That's the signal that the design has done its job on you. The reward was never the point. The two taps were.

If you're on the product side, the honest question is whether your tier structure would survive being shown to a customer without the progress bar. If the answer is no — if the value only exists in the framing — you've built something that works right up until the moment people understand it. And they always, eventually, understand it. The programs that last are the ones where the 90% is real, and the two taps are worth taking on their own merits, not because a bar told you they were.